Discovery-Driven Digital Transformation Success

By: Stephen Pappas

There's a graveyard most businesspeople don’t talk about. Not because it's not right there, but nobody sends out any invitations.

It's full of digital transformation projects that cost millions, launched with a “Go-Live” party, and then quietly went away about six months later when the adoption plummeted. Portals that nobody logged into anymore. Self-service platforms that just forced frustrated customers to head right back to the contact center for help. AI-powered tools, the frontline team found so many creative ways around.

I've sat through enough project postmortems to know how it goes. There's almost always a point – usually late in the meeting, after everyone's good and exhausted – where someone says it quietly: "We thought we knew what they needed."

That statement right there? That's your actual cause of potential project death.

Now, let’s talk a little about the reality: every single one of those failed projects had a discovery phase. One person on the team had checked that R box. They scheduled and held workshops. They sent out employee surveys. They most likely gathered stakeholders for journey mapping sessions with the color-coded sticky notes covering an entire conference room wall.

The discovery happened. It just wasn't the right kind of discovery or yielded the right insights to build a plan on top of.

The data paradox

I cannot remember a time in business (in my career) where we've known more about our customers – and understood them less.

Think about what's sitting in your tech stack right now. Sentiment analysis. Net Promoter Scores. Session recordings. Heat maps. Forecast and churn predictions. Usage and adoption dashboards that update information in real time. If you printed it all out, you could bury a small town. And yet somehow, organizations are still launching digital products, portals, bots, and more that miss the mark so much so that it would be funny if the price tags weren't so high in the stratosphere.

Nobody wants to reveal this, but here it is: data tells you what happened. The context tells you what it meant.

Your analytics platform can tell you that 68 percent of users abandoned the onboarding flow at step four. What it can't tell you is why – whether they were confused, distracted, skeptical, felt something was wrong, or just got pulled into a meeting. That distinction isn't some minor detail. It's the entire ballgame. Because "redesigning step four" and "rebuilding trust earlier in the experience" are completely different problems. They also both carry a much different price tag, too.

Here is the uncomfortable truth: if your discovery process begins and ends with pulling a report, you don't really have a discovery process. You've got something closer to a confirmation engine – a sophisticated, expensive system for finding evidence that supports whatever you already knew. The dashboards aren’t exactly lying to you. It's more like they are only telling you half the story. And half the story, in digital transformation, is how you end up in that project graveyard.

What real discovery looks like

Real discovery is a conversation. It needs to be an honest conversation, it needs conversational patience, it needs the right questions, and a real willingness to hear something that potentially fouls up your current plan.

That is how I think about discovery in two separate layers.

Layer 1 is the stated problem – what stakeholders tell you when you ask them directly. It's usually coherent, reasonably accurate, but almost never the whole story. Customers will tell you the app is confusing or not very intuitive. Some employees could tell you the new system causes their work rhythm to slow them down. None of these answers are fundamentally wrong. They're just too surface-level. And surface-level is where most organizations stop discovering and move on.

Layer 2 is the actual problem – what surfaces when you go deeper, past the polished complaint to the raw problems, issues, or challenges that are underneath the stated problem. This is where discovery gets more uncomfortable. This is the place where discovery gets genuinely useful.

I watched a hospital system spend considerable time and money getting ready to overhaul its patient portal application because the patient engagement numbers were low. The data showed a usability problem. Layer 1 discovery confirmed it to be true – patients did find the portal confusing. But when we kept asking, “why”, a different story emerged.